Commit 5d317375 authored by Misbah Mubarak's avatar Misbah Mubarak

Adding units to the network model documentation

parent 4c07ff24
...@@ -57,7 +57,7 @@ model-net parameters and model-specific parameters. The general parameters are: ...@@ -57,7 +57,7 @@ model-net parameters and model-specific parameters. The general parameters are:
knowing the particular models used. All models used in the CODES knowing the particular models used. All models used in the CODES
configuration file must be referenced in the parameter. See configuration file must be referenced in the parameter. See
doc/example_heterogeneous in the codes-base repository. doc/example_heterogeneous in the codes-base repository.
* packet_size - the size of packets model-net will break messages into * packet_size - the size of packets in bytes. model-net will break messages into
* modelnet_scheduler - the algorithm for scheduling packets when multiple * modelnet_scheduler - the algorithm for scheduling packets when multiple
concurrent messages are being processed. Options can be found in the concurrent messages are being processed. Options can be found in the
SCHEDULER_TYPES macro in codes/model-net-sched.h). In particular, the SCHEDULER_TYPES macro in codes/model-net-sched.h). In particular, the
...@@ -125,15 +125,16 @@ performed in: ...@@ -125,15 +125,16 @@ performed in:
(PADS), 2014. (PADS), 2014.
The configuration parameters are as follows: The configuration parameters are as follows:
* n_dims - the dimensionality of the torus * n_dims - the number of torus dimensions.
* dim_length - the dimensions of the torus. For example, "4,2,2,2" describes a * dim_length - the length of each torus dimension. For example, "4,2,2,2" describes a
four-dimensional, 4x2x2x2 torus. four-dimensional, 4x2x2x2 torus.
* link_bandwidth - the bandwidth available per torus link * link_bandwidth - the bandwidth available per torus link (specified in GiB/sec).
* buffer_size - the buffer size available at each torus node * buffer_size - the buffer size available at each torus node. Buffer size is measured
in number of flits or chunks (flit/chunk size is configurable using chunk_size parameter).
* num_vc - number of virtual channels (currently unused - all traffic goes * num_vc - number of virtual channels (currently unused - all traffic goes
through a single channel) through a single channel).
* chunk_size - element size per transfer. Messages/packets are sent in * chunk_size - element size per transfer, specified in bytes. Messages/packets are sent in
individual chunks. This is typically a small number (e.g., 32 bytes) individual chunks. This is typically a small number (e.g., 32 bytes).
== Dragonfly == Dragonfly
......
...@@ -71,8 +71,8 @@ within the same group. ...@@ -71,8 +71,8 @@ within the same group.
two groups with each other. two groups with each other.
- cn_vc_size: Number of packet chunks (default: 32 bytes) that can fit in the channel connecting the network - cn_vc_size: Number of packet chunks (default: 32 bytes) that can fit in the channel connecting the network
node with its router. node with its router.
- local_bandwidth: bandwidth of the channels connecting the routers within the same group. - local_bandwidth: bandwidth of the channels in GiB/sec connecting the routers within the same group.
- global_bandwidth: bandwidth of the global channels connecting routers of two different groups. Note than - global_bandwidth: bandwidth of the global channels in GiB/sec connecting routers of two different groups. Note than
each router has 'h' number of global channels connected to it where a=2p=2h in our configuration. each router has 'h' number of global channels connected to it where a=2p=2h in our configuration.
- cn_bandwidth: bandwidth of the channel connecing the compute node with the router. - cn_bandwidth: bandwidth of the channel connecing the compute node with the router.
** All the above bandwidth parameters are in Gigabytes/sec. ** All the above bandwidth parameters are in Gigabytes/sec.
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment